Marantz 2226B blows AC fuse

EastPoint

Recycler
Subscriber
Hi folks. I just picked up a 2226B that pops the rear AC fuse. I replaced the fuse and powered it up on a dim-bulb tester and it lights the bulb nice and bright. A friend thinks it’s a bad diode bridge. Is it likely to be that, or something else? I’ll start digging into it soon but I figured I’d get some opinions first. I didn’t see any similar threads from the past.
 
Shorted output(s) would be my first guess then bad main caps.
I have rarely seen a full wave bridge rectifier assembly go bad.

Bob
 
I pulled two of the outputs (left side of the heat-sink) and both were shorted. I powered it up on the dim-bulb with those two out and it didn't light the bulb. Resistor R756 (120 ohm 1/2w) on that side started smoking so I shut it off. I'm guessing I should replace all four output transistors, all four drivers, and that resistor, plus the four glass-bead diodes. I'll get the parts ordered and go from there.

I found an old thread that said I should use MJE15031 and MJE15030 to replace the outputs. I believe I'm supposed to replace the 2SA777 driver transistors with KSA1220s as well, and I'll replace the 2SA733 transistors with KSA992s while I'm at it.

Do I swap all the glass bead diodes (green with black on the end) with a pair of 4148 diodes in series each?
 
Last edited:
For the diodes I assume you mean Q705-708? Those are MV-13’s according to the SM and should be replaced with 2 4148’s in series as you mentioned. I almost always hear them referred to as epoxy diodes hence my clarification. Not sure which term is more accurate.
 
Yes, the MV-13s. I always call them "glass bead diodes." I swapped those out just now and put in the KSA992s, but I'll wait til the outputs get here before I do more.
 
Alright, so I replaced the outputs and drivers on the bad channel. Specifically, I replaced:
All epoxy diodes with paired 4148s in series
Q734 is now MJE15030G
Q736 is MJE15031G
Q726 is KSC2383
Q728 is KSA1013
Q717 and Q718 are KSA992
Q712 and Q714 are 4148 diodes
replaced R756 with a 120 ohm 1/2W resistor

It came up and played fine for a second but the new outputs were getting REALLY hot. I powered it off and used the metal bar to push them up against the heat sink. When I powered it up again I was getting heavy crackling (so presumably spitting voltage) on that channel. The other channel still works fine. I didn't adjust the bias or anything yet.
Thoughts?

IMG_0300.JPG IMG_0301.JPG IMG_0302.JPG IMG_0303.JPG
 
Last edited:
Make sure if there’s a metal tab on the mje15030 that it’s completely insulated from the heat sink. There should be a full mica insulator from the factory. It’s probably ok to reuse.
 
Right, they're not attached to the heat-sink. They're normally insulated from it anyway, so it shouldn't be a grounding issue.
 
Right, they're not attached to the heat-sink. They're normally insulated from it anyway, so it shouldn't be a grounding issue.
will be a heat issue though .. number 1 rule . never run transistors that mount on a heat sink without bolting them to a heat sink ..
 
yes it can do .. however if the idle current was dialled right back i guess they could sit a long time whilst they are turned off . no point even doing that though .
 
you should be using a dbt for starters . then see how the idle bias is looking with good transistors installed . it needs dialling down by the looks of it . maybe there is a fault in the bias circuit that blew it up in the first place .
 
Back
Top Bottom